The third and final canon of Greek majuscule is today generally referred to as "ogival", a name which simply refers to the "oval" shape of many of its letters; in earlier scholarship it was often called "Slavic" (as we saw earlier with the "Coptic" majuscule, here again the ethnic designation is now considered inappropriate because the Cyrillic script is in fact derived from a variant of this one, not vice-versa). It is characterized by:
- strong, binary stroke contrast, with vertical and descending strokes tending to be thick and the others thin;
- broken curves, often at acute angles, especially at the top and base lines;
- the oval shape (tending in fact to a diamond shape, due to the previous characteristic) of the four letters which, in the Biblical majuscule, are circular (epsilon, theta, omikron, sigma);
- a bilinear scheme which is broken by the descenders of rho and of ypsilon, the ascenders and descenders of phi and psi, and often also by the lower parts of zeta and xi, as well as by the lower left stroke, or both lower strokes, of chi;
- a tendency to use triangular serifs, especially at the extremities of horizontal strokes like the cross bar of tau, that of theta (which extends beyond the rounded part of the letter on either side), the lower stroke of delta (which extends beyond the two oblique strokes on either side), and the right end of the cross-bar of gamma or of the lowest strokes of zeta and xi. Serifs are not used when the extremity of the horizontal stroke joins another stroke (such as the cross bars of pi and eta, which do not extend beyond the vertical strokes).
This last feature, which actually appears to be a relatively late innovation (perhaps from the seventh c. onwards) rather than an essential component of the canon, greatly contributes to its "Slavic" appearance, which gave rise to its earlier name.
This script appears in two variants, one with a vertical axis (known as "upright ogival majuscule") and one with a slanted axis ("slanted ogival majuscule"). Both may be observed together in the eighth- or ninth-century Vat. gr. 749 (seen below), where the Greek text of the book of Job is in slanted ogival majuscule, while the marginal commentary (or "catena", i.e. "chain" of comments culled from patristic authors) is in upright ogival majuscule.
Vat. gr. 749 pt. 2 (initial view: f. 122v, Job 19:12-13)
In older scholarship you will find the idea that the slanted type was the earlier one and was supplanted by the upright type around the fifth century CE. In fact they were used contemporaneously, as illustrated in the image above; it is the dearth of dated examples which has given rise to this and other questionable theories. A more likely view is the one put forward by E. Crisci in an important article published in 1985 ("La Maiuscola ogivale diritta: origini, tipologie, dislocazioni", Scrittura e civiltà 9, pp. 103-145), namely that both types developed in parallel out of scripts like the one we see in the famous second-century Bacchylides Papyrus (the idea had already been suggested by Cavallo, "Note sulla scrittura greca libraria dei papiri", Scriptorium 26 , p. 73 = Il Calamo …, p. 103). The question of origins is in any case purely theoretical; a more practical question is whether dates (or geographical origins) may in fact be credibly assigned to hands using this script. Both Cavallo (in Ricerche, pp. 117-121, writing specifically about the slanted type) and Crisci (writing about the upright type, in the cited article, p. 120) hypothesized a development which may be summarized as proceeding from an early period with less stroke contrast and with few or no serifs to a late "mannerism" with increasingly prevalent use of such features, esp. from the seventh century onwards. As with Cavallo's reconstruction of the development of the Biblical majuscule, this proposal is not implausible, but it is difficult to support it with actual data, since there are very few dated or securely datable manuscripts written in this hand (lists of these are given in Crisci's article, pp. 117-118, and in P. Orsini, "La maiuscola ogivale inclinata. Contributo preliminare", Scripta 9 , pp. 89-116:96-98; 104-105). Certainly a comparison of Vat. gr. 1291 (securely datable to the period between 813 and 820 CE) and Vat. gr. 354 (dated to 949 CE) shows greater "mannerism" in the later script than in the earlier one, as you can see below.
(left) Vat. gr. 1291 (813-820 CE) (initial view: f. 91v); (right) Vat. gr. 354 (949 CE) (initial view: f. 24v).
However, the differences between these two scripts could simply reflect the personal styles of these two individual scribes, or be due to any number of other factors, such as the fact that the former is a collection of astronomical tables while the latter is a literary text; and perhaps also the fact that the former manuscript was almost certainly copied in Constantinople, while the latter one is likely of "provincial" origin. There have in fact been attempts to assign geographical origins to manuscripts written in ogival minuscule, which have mostly concerned the slanted variant of the script, and have been based mostly on observations about the degree of the slant (with the westernmost scribes having — hypothetically — the least slant and the easternmost ones the greatest); but these attempts have not reached convincing results either, as detailed by Orsini in the already mentioned article, pp. 112-116. In fact, as Orsini rather drily observes (p. 93), the whole idea of placing scripts on the preconceived developmental arc of a "canon," necessarily running "dalle semplicità delle origini ad una sempre maggiore complessità ed artificiosità," which is implicit in Cavallo's and Crisci's approaches, is problematic, as has in fact been recently recognized by Cavallo himself (La Scrittura greca e latina dei papiri, p. 15) and, implicitly at least, by Crisci (in the book he co-authored with P. Degni, La scrittura greca dall'antichità all'epoca della stampa, pp. 112-118). Still, as for the Biblical majuscule, so for the ogival one, in the absence of reliable data, the development proposed by Cavallo and Crisci may at least be taken as a plausible rule of thumb: in other words, generally speaking, scripts with preponderant serifs and a marked contrast between thick and thin strokes may be taken to be later than ones which lack these features.
For practice reading and transcribing this script, please see Vat. gr. 2066